Search

Several approaches and techniques exist to resolve load balancing problem in general and torus topology networks. Graph methods are natural ways to perform balancing of routing paths. A routing balancing algorithm must operate within the constraints of the underlying network architecture that limits several parameters, such as the number of logical paths in the network. In this paper, we consider a torus topology that is one of the common topologies that are used in high performance computing systems. We introduce a routing graph that corresponds to a interconnect topology, abstracts interconnect routing rules, providing one-to-one correspondence of network routes and graph paths. We propose a deadlock-free routing algorithm based on a fast single-source shortest path algorithm in a routing graph for the deterministic routing of a torus topology interconnect with a single virtual channel. The proposed algorithm is a tradeoff between two existing routing algorithms, also the algorithm is optimized for multidimensional torus topology and can be applied to any torus topology HPC system. We present a complete description of a routing graph that abstracts the Angara interconnect with 4D torus topology. We evaluated the proposed routing algorithm and benchmarked the maximum performance improvement of 71% for the Alltoall pattern for torus topology systems up to 432 nodes on the Angara interconnect simulator. The performance improvement of more than 5% was obtained for the NPB FT and IS application kernels on a 32-node supercomputer. • A routing graph abstracts network routing rules. • A deadlock-free routing algorithm is based on a fast single-source shortest path algorithm and optimized for a torus topology. • A routing graph for the Angara network reflects the restrictions on a further packet route with the previous route steps in mind. • The maximum performance improvement is 71% for the Alltoall pattern on the simulated torus topology systems up to 432 nodes. • The performance improvement on the 32-node Angara based cluster is 5% for the NPB FT and IS application kernels. [ABSTRACT FROM AUTHOR]

This paper provides an in-depth understanding of social dynamics in the form of kinship ties in matrilineal societies. It unpacks gender roles and relationships at the community level to understand how social structures, created by the pattern of relations, enhance or hinder coping initiatives during lake recessions in the Lake Chilwa socio-ecological system. Using qualitative and quantitative methods including in-depth interviews, Focus Group Discussions (FGD) and household questionnaires, this paper provides insights into how people in the Lake Chilwa social-ecological system prepare for and cope with Lake Chilwa water recessions by capitalising on their social networks. The findings have shown that during lake recessions poor households in the Lake Chilwa socio-ecological system are cushioned by fellow households through lineage networks. There is strong tradition secured through kinship ties where the generic term ‘mwambo wathu’ (our tradition) embraces a group of formative norms enforced via a series of rules and rituals. Based in a matrilineal system, women have strong rights over land allocation and use in which the female sorority group ‘mbumba’ is very stable. There is a great deal of reciprocity and food redistribution among the households. This self-organization provides deeper meaning about how people assimilate the dynamics affecting their social and natural environments. Specifically the findings provide evidence on how households, communities and their livelihoods in Africa are reconstructed as natural resources fluctuate. [ABSTRACT FROM AUTHOR]

Abstract Geometric gradient flows for elastic energies of Willmore type play an important role in mathematics and in many applications. The evolution of elastic curves has been studied in detail both for closed as well as for open curves. Although elastic flows for networks also have many interesting features, they have not been studied so far from the point of view of mathematical analysis. So far it was not even clear what are appropriate boundary conditions at junctions. In this paper we give a well-posedness result for Willmore flow of networks in different geometric settings and hence lay a foundation for further mathematical analysis. A main point in the proof is to check whether different proposed boundary conditions lead to a well posed problem. In this context one has to check the Lopatinskii–Shapiro condition in order to apply the Solonnikov theory for linear parabolic systems in Hölder spaces which is needed in a fixed point argument. We also show that the solution we get is unique in a purely geometric sense. [ABSTRACT FROM AUTHOR]

Abstract: This paper analyzes the impact of infrastructure investment on Spanish economic growth between 1850 and 1935. Using new infrastructure data and VAR techniques, this paper shows that the growth impact of local-scope infrastructure investment was positive, but returns to investment in large nation-wide networks were not significantly different from zero. Two complementary explanations are suggested for the last result. On the one hand, public intervention and the application of non-efficiency investment criteria were very intense in large network construction. On the other hand, returns to new investment in large networks might have decreased dramatically once the basic links were constructed. [Copyright &y& Elsevier]

This paper describes and validates a novel framework using the Approximate Bayesian Computation (ABC) algorithm for parameter estimation and model selection in models of mesoscale brain network activity. We provide a proof of principle, first pass validation of this framework using a set of neural mass models of the cortico-basal ganglia thalamic circuit inverted upon spectral features from experimental, in vivo recordings. This optimization scheme relaxes an assumption of fixed-form posteriors (i.e. the Laplace approximation) taken in previous approaches to inverse modelling of spectral features. This enables the exploration of model dynamics beyond that approximated from local linearity assumptions and so fit to explicit, numerical solutions of the underlying non-linear system of equations. In this first paper, we establish a face validation of the optimization procedures in terms of: (i) the ability to approximate posterior densities over parameters that are plausible given the known causes of the data; (ii) the ability of the model comparison procedures to yield posterior model probabilities that can identify the model structure known to generate the data; and (iii) the robustness of these procedures to local minima in the face of different starting conditions. Finally, as an illustrative application we show (iv) that model comparison can yield plausible conclusions given the known neurobiology of the cortico-basal ganglia-thalamic circuit in Parkinsonism. These results lay the groundwork for future studies utilizing highly nonlinear or brittle models that can explain time dependant dynamics, such as oscillatory bursts, in terms of the underlying neural circuits. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, we propose a discrete version of the following semilinear heat equation with absorption with , which is said to be the ω-heat equation with absorption on a network. Using the discrete Laplacian operator on a weighted graph, we define the ω-heat equations with absorption on networks and give their physical interpretations. The main concern is to investigate the large time behaviors of nontrivial solutions of the equations whose initial data are nonnegative and the boundary data vanish. It is proved that the asymptotic behaviors of the solutions as t tends to +∞ strongly depend on the sign of . [Copyright &y& Elsevier]

Abstract: Rheology of sepiolite-based epoxy suspensions as well as morphology and dynamic mechanical properties of the corresponding nanocomposites are discussed in this paper. The influence of the type of sepiolite used, i.e. non-modified, aminosilane and glycidylsilane surface modified, and of the process developed to prepare the epoxy suspensions were investigated. Except for low amount of filler, a shear thinning behavior was observed in the others sepiolite-based epoxy suspensions. The interactions developed between the sepiolite and the epoxy matrix are responsible for the magnitude of the shear thinning effect and are related to the morphology of the nanocomposites. The best dispersion of sepiolite was achieved using either an emulsion process or a glycidyl functionalized sepiolite. [Copyright &y& Elsevier]

Abstract: This paper describes the parallel implementation of a transport network model. ‘Single-Program, Multiple Data’ (SPMD) paradigm is employed using a simple data decomposition approach where each processor runs the same program but acts on a different subset of the data. The objective is to reduce the execution time of the model. The computationally intensive part of the model is within the assignment and simulation section and therefore this section is parallelised and executed using 1, 2, 4, 8 and 16 processors. The convergence, accuracy and performance of the parallel model are then assessed and compared to the linear implementation. The results indicate a performance increase of over 8 for the parallelised module and a speedup of 5 for the total model when the model is run using 16 processors. The efficiency, average parallelism and efficiency-execution time profile are also discussed. In the context of time savings with 16 processors compared with 1, the time saving on the IBM SP2 are of the order of 80%, and, compared to a linear implementation on a dual processor Intel machine are of the order of 86%. [Copyright &y& Elsevier]

Line of Sight (LoS) networks were designed to model wireless communication in settings which may contain obstacles. For fixed positive integer d , and positive integer ω , a graph G = (V , E) is a (d -dimensional) LoS network with range parameter ω if it can be embedded in a finite cube of the d -dimensional integer grid so that each pair of vertices in V are adjacent if and only if their embedding coordinates differ only in one position and such difference is less than ω. In this paper we investigate a dynamic programming (DP) approach which can be used to obtain efficient algorithmic solutions for various combinatorial problems in LoS networks. In particular DP solves the Maximum Independent Set (MIS) problem in LoS networks optimally, for any ω , on narrow LoS networks (i.e. networks which can be embedded in a k × k ... × k × n region, for some fixed k independent of n). In the unrestricted case it has been shown that the (decision version of the) problem is NP-hard when 2 < ω = O (n 1 − δ) , for fixed 0 < δ < 1. We describe how DP can be used as a building block in the design of good approximation algorithms in this case. In particular we present a semi-online polynomial-time approximation scheme for the MIS problem in narrow d -dimensional LoS networks, as well as a polynomial-time 2-approximation algorithm and a fast polynomial time approximation scheme for the MIS problem in arbitrary d -dimensional LoS networks. Finally we comment on how the approach can be adapted to prove similar results for a number of important optimization problems in LoS networks. [ABSTRACT FROM AUTHOR]